Isn't that graph hopelessly tainted by eg. the dominance of WordPress? I mean, yes that's PHP, but just because 80% of sites are running it doesn't mean 80% of development is in PHP. Far from it.

All those kind of charts have the same problem. It’s like asking “what food do people like the most?”, and answering “rice” because so many people eat it every day. The problem is that the question is so vague you can answer it many different ways, and all of the answers are correct.

What people usually mean when they use those stats is “look how popular my choice of language is”.

It is, but discounting PHP because of WordPress from that graph is still interesting - I doubt there are free one-click installs for ASP.Net like there are for WordPress for example, but are there for Ruby on Rails?

The comparison does not make sense. WordPress is a CMS that happens to be written in PHP, while ASP.NET and Ruby on Rails are frameworks for building web services which don't do anything by themselves. At best you would get a blank "Hello world" page and nothing more from a "one-click" installer.

That graph counts site using Shopify. And shopify represent 80%+ of RoR usage.

Yeah, three companies worth an estimated $300B can keep any language alive. It’s fascinating to track the second wind of out-of-fashion languages, held aloft by a few mature companies that depend on it.

I'd argue even without them Ruby would be alive in the sense that the core language would get frequent updates. Even Perl is still alive https://github.com/Perl/perl5, even in a healthier state than I would have thought.

Oh wow, Perl is still getting monthly updates - I knew it was on the decline but I thought the whole Perl 6/Perl 7/Perl 11 fragmentation had left things in a worse state.

I’ve heard of Raku (Perl 6), but didn’t realize there was a Perl 7 or Perl 11. Pretty interesting problem. In Java land you get Java 8, current Java, Groovy, Kotlin, etc. So Perl is in good company in this regard.

Perl 11 was a thought experiment. It never actually became anything beyond a website expressing that notion. It got no traction from anybody apart of the originators. And of that I'm not really sure either.

Facebook and PHP are kind an interesting twist on that. Facebook spent a lot of time and money on Hack/HHVM, then the PHP core devs responded with a new version of PHP that is, in many cases, faster. Though Hack/HHVM is still better for async work.

Gitlab, too